NSCT Transform Based Medical Image Spatial Frequency Excitation PCNN Fusion Algorithm

Image fusion is a specific algorithm, which using two or more images to create a new image. The main idea is that:by the type of imaging mechanism, the type of the operating wavelength, and a variety of sensors of the type of working environment for the same scene, the picture of the plurality of the imaging information fusion as a new image. And it combine with a certain algorithm on the basis of the amount of information to get higher reliability, lower degree of blurring, better understandability, more consistent with human visual requirements. Image fusion is suitable for the computer to detect, identify, classify, and understand or other operation processing.And medical image fusion is a branch in the image fusion field, it is closely related to people’s life, and it has very important practical significance:the medical image of the same parts in the human body but take via different sensors, by using image fusion technology to fuse the complementary information to be a new medical image, so this is a picture with higher quality, more clear, more complete medical image, so doctors can better analysis or diagnosis patient condition. By this method, doctors will be able to effectively avoid the risk of misdiagnosis. Therefore, on account of this concept, this paper puts forward a medical image fusion algorithm, which based on NSCT transform combining with spatial frequency transform under the incentive of PCNN algorithm.CT and MRI images are types of image with different imaging principle optical image picture for brain tissue structure. Via fusion operation of CT image and MRI image, you can get the soft tissue and bone tissue are clear and fully reflect the condition of the fused image, to provide more comprehensive clinical, more effective, more direct, more reliable information, thereby improving the clinical diagnosis efficiency, reliability and accuracy is greatly improved, and greater than each of the partial information of the final result so that the overall information of the type of medical image.The image registration prior to the fusion operation is an essential step. It can avoided the relative translation, rotation and scaling difference due to the different imaging device imaging orientation, angle and resolution type, etc, or by other factors in collected medical image. The paper is based on Powell algorithm and presented an image registration method based on multi-resolution and improved Powell algorithm. A better solution is focused on the function into a local extreme, complex structure, not easy programming and other issues; Image fusion algorithm based on NSCT transform combined with spatial frequency under the incentive of PCNN algorithm of medical image fusion algorithm, by made of made a contrast experiment, the results show that the method put forward by the paper effect significantly improved.